Further Notes on the Port Kennedy Cave Arvicolid Rodents

Paludicola 9(1):13-20 November 2012 © by the Rochester Institute of Vertebrate Paleontology

FURTHER NOTES ON THE PORT KENNEDY CAVE ARVICOLID RODENTS

Robert A. Martin

Department of Biological Sciences, Murray State University, Murray, KY 42071

ABSTRACT Cope (1871) named five Microtus-like species from the Port Kennedy Cave deposit of Montgomery County, Pennsylvania, all of which he referred to the genus Arvicola. Using Cope’s illustrations and descriptions, a review by C. W. Hibbard and examination of three Port Kennedy specimens, all of Cope’s species (Arvicola speothen, A. tetradelta. A dideltus, A. sigmodus and A. involutus) are deemed nomina dubia. Two available specimens are clearly referable to Pitymys cumberlandensis van der Meulen, and another is referred to cf. Microtus (Pedomys) llanensis Hibbard. On the basis of these specimens and other considerations, Port Kennedy is considered to be younger than the Cumberland Cave local fauna of Maryland and approximately contemporaneous with the Trout Cave 2 local fauna of West Virginia, falling within Great Plains Rodent Zone (RZ) 14, between about 1.3 - 0.64 Ma. The origin of Pitymys is obscure. Absence of a suitable North American ancestor and dental features of P. cumberlandensis suggest we should look for an Old World immigrant ancestor with undifferentiated enamel, a complex m3, and an m1 with five closed triangles and deep, provergent reentrants. Confluent T1-2 on lower molars and T2-3 on M3 in P. cumberlandensis may preclude this species as an ancestor for later North American pine voles.

INTRODUCTION Hibbard (1955) valiantly attempted to make sense of the material from Cope’s (1871) descriptions and Six species of arvicolid rodents were described by comparisons made by Gidley and Gazin (1938) when Cope (1871) from the Port Kennedy Cave deposit in they described arvicolid material from the Cumberland Montgomery County, Pennsylvania. All were referred Cave deposit of Maryland. Van der Meulen (1978) to the genus Arvicola. The taxonomic history of these later restudied the arvicolids from the Cumberland species was reviewed by Hibbard (1955), and he Cave deposit of Maryland, referring them to Microtus synonymized five of the Microtus-like species under paroperarius Hibbard (1944) and the new species Microtus speothen, “Pedomys or Pitymys “ dideltus Pedomys guildayi and Pitymys cumberlandensis. In his and “Pedomys or Pitymys” involutus. The sixth paper, van der Meulen (1978) considered “Pitymys” Hibbard identified as an extinct muskrat, Ondatra involutus as a nomen dubium, but did not consider hiatidens, which is probably juvenile O. annectens Cope’s (1871) other species. From Hibbard’s (1955) (Martin et al., 2009). He also added the genus Neofiber descriptions and illustrations of the Port Kennedy to the arvicolid community. Hibbard examined material material, van der Meulen’s (1978) classic work on the in the American Museum of Natural History (AMNH) Cumberland Cave arvicolids and subsequent research and a few specimens from the Academy of Natural on arvicolids from northeastern states (Pfaff, 1990) it Sciences of Philadelphia (ANSP), as the Port Kennedy seems likely that the Port Kennedy Cave arvicolid specimens were split between the two institutions. The cadre includes a species of Microtus near Hibbard’s holotypes were kept at the AMNH and, according to (1944) M. paroperarius and two additional species Hibbard (1955) were in the following condition when near van der Meulen’s (1978) Pitymys cumber- he examined them: 1) Arvicola speothen (AMNH landensis and either van der Meulen’s (1978) Pedomys 8689) – imprint of a left ramus with only m3 guildayi or Hibbard’s (1944) Pedomys llanensis. The preserved; 2) Arvicola tetradelta (AMNH 8692) – first question that arises is: Are any of Cope’s according to Cope (1871) is a maxillary piece with M2- holotypes or topotypes sufficiently preserved to be M3 but Hibbard (1955) identified the molars as m2- diagnostic at the species level, perhaps supplanting one m3; 3) Arvicola involuta (AMNH 8699a) – teeth or more of Hibbard’s and van der Meulen’s later absent, presumably disintegrated. 4) Arvicola sigmodus names? Combining information from Cope’s (1871) (AMNH 8696) – molars missing, only piece of ramus original work, Hibbard’s (1955) review and personal and lower incisor remains; 5) Arvicola didelta (AMNH obervation of three specimens, I review the Port 8694) – all teeth missing, presumably disintregrated. Kennedy Microtus-like voles and tentatively correlate the Port Kennedy Cave assemblage with other early

14 MARTIN—PORT KENNEDY ARVICOLID RODENTS

Pleistocene assemblages from the eastern and the extinct Allophaiomys, are generally indistin- midwestern states. guishable, and as such it would not be possible (at least currently) to distinguish M. paroperarius from P. METHODS guildayi based on this molar. Relying on Cope’s (1871) description of the m1-m2 of AMNH 8689 before it was Upper molars are abbreviated by upper case lost, Hibbard (1955, p. 90) further referred the letters (e.g., M3), lowers by lower case letters (e.g., ambiguous AMNH 8692, the holotype of Cope’s m3). Three arvicolid specimens from the Port Kennedy Arvicola tetradelta, to M. speothen. However, even if Cave deposit housed at the Philadelphia Academy of the two molars of AMNH 8692 are eventually Natural Sciences (ANSP) were examined and identified correctly, they cannot be used to identify the measured for this study: ANSP 141 (figured by subgenus Microtus, let alone M. speothen or M. Hibbard, 1955; fig. 2D): part right (R) mandible with paroperarius. In the absence of additonal material that incisor and m1-m3; ANSP 142 (figured by Hibbard, could be construed to be part of a “type series,” I 1955; fig. 2B): part left (L) mandible with m1; ANSP recommend we consider Arvicola speothen and A. 18907: part L mandible with m1-m3. tetradelta as nomena dubia. Measurements were made with an AO filar ______micrometer coupled to an AO binocular microscope. The micrometer was calibrated with a 2.0 mm AO slide and measurements were multiplied by the appropriate correction factor (0.615 in this case). Measurements made on m1 are shown in Figure 1. Ratios calculated from the measurements are as follows: A/L = 100a/L, a measure of the relative length of the ACC; W’/W = 100w’/W, a measure of the relative width of the anterior cap; B/W = 100b/W, a measure of the degree of confluency between the anterior cap and the triangles 4 and 5; C/W = 100c/W, a measure of the degree of confluency between triangles 4 and 5 (Martin, 1991). I follow van der Meulen (1978) in assuming that T1 is not expressed on M2-M3; consequently the triangles begin on these teeth after the anterior loop with T2 instead of T1 as illustrated by Pfaff (1990; fig. 3). Dental abbreviations in the text are as follows: BRA = buccal re-entrant angle, LRA = lingual re- entrant angle. Figure 2 was traced from a digital photograph, reduced and darkened. Parts of some of the triangle edges were obscured by thick glue and are therefore best estimates. Nevertheless, most triangles clearly displayed undifferentiated enamel.

SYSTEMATIC PALEONTOLOGY FIGURE 1. Topography of a Microtus left m1 (after Pfaff 1990). ac FAMILY ARVICOLIDAE = anterior cap (sometimes also referred to as the acd, or anteroconid); COPE’S SPECIES L = occlusal length; a = length of ACC (anteroconid complex; all structures anterior to T3); W = greatest width ACC; w’= greatest Arvicola speothen Cope, 1871 width of anterior cap; b = width of dentine isthmus connecting T4- Arvicola tetradelta Cope, 1871 T5 and the anterior cap; c = width of dentine isthmus connecting T4 - T5; T1, T2 = triangles 1 and 2 (to identify other triangles, continue Comments—Despite the presence only of m3 counting from T2; buds from the ac are considered “incipient” triangles). (m1 and m2 are missing), Hibbard (1955) retained ______AMNH 8689 as the holotype of Microtus speothen and, without explanation, suggested that M. speothen and Arvicola involuta Cope, 1871 M. paroperarius from the Cudahy and Subrite local faunas of Kansas (Hibbard, 1944) might be Comments—Molars of the holotype of Arvicola conspecific. However, the m3 of most Microtus, involuta (AMNH 8694) have disintegrated and the including the extant subgenera Microtus, Pedomys and illustration of an m1 provided by Cope (1871, fig. 16) PALUDICOLA, VOL. 9, NO. 1, 2012 15

is not definitive of any Microtus-like subgenus or 15, suggest didelta is a member of the taxon Pedomys species, and it does not resemble the m1 of extant (Martin, 1991; Pfaff, 1990). However, both the series Pitymys pinetorum, despite Cope’s suggestion to the of isolated m1s and the holotype of A. didelta are contrary (Cope, 1871, p. 90). I recommend we consider missing. The topotype m1-m3 illustrated by Hibbard Arvicola involuta as a nomen dubium. (1955) is not advanced enough in dental anatomy to be ______distinguishable from Great Plains Pleistocene Microtus referred to M. (Allophaiomys) pliocaenicus (Martin, 1989). Although the molars of AMNH 8688 display positive enamel differentiation (unlike Pitymys cumberlandensis, in which the enamel is undifferentiated) both M. pliocaenicus and M. (Pedomys) guildayi from Cumberland Cave display positive enamel differentiation (van der Meulen, 1978; Martin, 1989, 1995). Because the remaining material of Arvicola didelta is not sufficient to define a distinct Microtus species, I recommend that A. didelta be considered a nomen dubium.

Arvicola sigmodus Cope, 1871

Comments—The molars from the holotype mandible of A. sigmodus, AMNH 8606, are missing, and none of the uncatalogued individual molars illustrated by Cope (1871, fig. 17) are currently recognized in the AMNH collections (R. Oleary, personal commmunication). Cope’s (1871) diagnosis of A. sigmodus is contradictory, indicating some similarity to species we currently recognize as Microtus (Pedomys) ochrogaster and Pitymys pinetorum, which is not surprising because he illustrated a combination of species under A. sigmodus in his fig. 17, one of which, fig. 17a, may not even be a Microtus or Pitymys. Fig. 17b of A. sigmodus is roughly similar to the m1 of a Pitymys species, but as it is missing we cannot be sure. I recommend we consider A. sigmodus as a nomen dubium.

MICROTUS-LIKE SPECIES FROM PORT KENNEDY CAVE

Microtus (Pedomys) cf. llanensis FIGURE 2. Occlusal morphology of Lm1-3 of Pitymys cumberlandensis (ANSP 18907) from Port Kennedy Cave. Scale bar Referred Material—ANSP 142, left mandibular = 1.0 mm. ______fragment with juvenile m1 (illustrated by Hibbard, 1955). Arvicola didelta Cope, 1871 Comments--As documented by Martin (1987) and confirmed statistically by Pfaff (1990), the subgenus Microtus dideltus (Cope, 1896) Pedomys can be recognized by a relatively shallow and Pedomys or Pitymys dideltus (Hibbard, 1955) C-shaped BRA3 (the anterior enamel border of T4 is Comments—Illustrations of three m1s and an relatively horizontal [rather than provergent] compared M3 in an M1-M3 series of Arvicola didelta (Cope, to this border in Pitymys). This feature is observable in 1871; fig. 15) are similar to the same teeth of Pedomys ANSP 142 (Hibbard, 1955, fig. 2B and personal guildayi van der Meulen (1978) based on occlusal obervation). The relatively thin enamel and acute morphology. The shallow and C-shaped BRA3 on m1, enamel borders identify ANSP 142 as a juvenile, and clearly distinguishable in Cope’s illustration in his fig. thus measurements and ratios of those measurements 16 MARTIN—PORT KENNEDY ARVICOLID RODENTS

(after van der Meulen, 1978) are needed to at least shallow and rarely contain cementum, complex m3 tentatively determine the identity of this specimen. with well-developed T4, and 3) deep penetration of Measurements and ratios from ANSP 142 are presented BRA3 and LRA3, almost closing T4 from T5. North in Table 1. In measurements and ratios ANSP 142 is American Pitymys can be identified by a prevalence of indistinguishable from both M. llanensis and M. m1s with deep and provergent BRA2, BRA3 and guildayi. Nevertheless, the relatively advanced form of LRA3 and LRA4. Deep penetration of BRA3 into the the anteroconid on ANSP 142 as a juvenile suggests acd often isolates T6 as a sausage or finger shape with that it would have more approximated average M. the apex pointing posteriorly (Figure 2). ANSP 141 llanensis in its adult configuration than average M. (illustrated by Hibbard, 1955) and ANSP 18907 guildayi, and is tentatively referred to M. llanensis. (Figure 2) display the salient features of P. In a review of fossil Pedomys, Martin (1995) cumberlandensis, although T4 is somewhat reduced on synonymized Pedomys guildayi (van der Meulen, m3 in ANSP 141. Van der Muelen (1978) did not 1978) under M. pliocaenicus and M. llanensis provide description of variation in the m3 of P. (Hibbard, 1944) under M. ochrogaster to reflect the cumberlandensis, and for now the differences between likelihood, first expressed by van der Meulen (1978), ANSP 141 and 18907 are considered to represent that the extant M. ochrogaster evolved by phyletic population variation. evolution from early Pleistocene Great Plains The B/W measurement for ANSP 18907 (Table populations referred to Microtus pliocaenicus (Martin, 1), representing the relative width of the dentine 1989). Whereas M. ochrogaster certainly evolved from isthmus connecting T5 with the anteroconid and a species similar to M. pliocaenicus of the Java local therefore the tendency for T5 to be closed from the acd, fauna from South Dakota, the recognition of two new is outside the small end of the range for B/W in P. extinct Pedomys species, M. australis from Florida and cumberlandensis from Trout Cave 2, implying a very Texas (Martin, 1995), and M. parmaleei from Kansas narrow connection. This condition is certainly derived and the Appalachian chain in the eastern United States in North American voles as compared with early (Martin, 1991; Martin et al., 2012) demonstrates that a Microtus (Allophaiomys) from the Java local fauna single species phyletic model is no longer tenable. (Martin et al. 1998), but it appears to have evolved Because some specimens of M. guildayi demonstrate independently a number of times in unrelated voles, the shallow, C-shaped BRA3 of Pedomys, it now including other Pitymys (e.g., P. aratai, M. appears more prudent to resurrect both M. guildayi and guatemalensis; Martin, 1987,1995). Currently, this M. llanensis, and then to try to determine their condition in ANSP 18907 is considered to represent an relationships to other extinct and extant species of the extreme in P. cumberlandensis population variation, prairie vole clade with a more complete phylogenetic but without taxonomic significance. analysis. Pedomys is here retained within Microtus Van der Meulen (1978) was impressed by the until an analysis demonstrates that early Pleistocene tendency in some extant P. pinetorum to retain North American Microtus cf. pliocaenicus molars can undifferentiated enamel, but this must be rare, as most be separated unequivocally from Eurasian Microtus of the specimens of extant P. pinetorum I have studied pliocaenicus molars. If they cannot, then it seems display positive differentiation, although enamel on the reasonable to assume that many extant Microtus in anterior triangle borders in Pitymys is distinctly thicker both the Old and New World with positive enamel than in Microtus at the same relative molar size differentiation are descended from the evolutionary (Martin, 1991: table 2). The m3 is also simpler (T4 queen species M. pliocaenicus (Martin, 1995). This reduced) in extant P. pinetorum then it is in P. was essentially the conclusion reached earlier by van cumberlandensis. The shift to positive enamel der Meulen (1978, 114). differentiation and change in form of m3 are dental changes that characterize the evolution of later Pitymys Pitymys cumberlandensis van der Meulen, 1978 species as they branched from P. cumberlandensis or a related species. In his diagnosis of P. cumberlandensis Pedomys or Pitymys involutus? (Hibbard), 1955 (fig. van der Meulen (1978) also included the form of M3, 2D) which he labeled as unreduced, in combination with the Referrred Material—ANSP 141, part right m3. Judging by his illustrations, I assume van der mandible with m1-m3; ANSP 18907, partial left Meulen was referring primarily to the relatively deep mandible with m1-m3. and postvergent LRA3, but this feature is highly Comments—Pitymys cumberlandensis (van der variable in extant P. pinetorum, as shown by Arata Meulen, 1978) is a medium-sized Pitymys (1965: fig. 4). M3s of extant P. pinetorum from the characterized by the following: 1) relatively thick, northeastern and midwestern United States display an undifferentiated enamel, 2) relatively simple equal complexity to those of P. cumberlandensis. Only anteroconid on m1 in which BRA4 and LRA5 are P. pinetorum parvulus from Florida displays reduced PALUDICOLA, VOL. 9, NO. 1, 2012 17

TABLE 1. Comparative measurements (in mm) of Pitymys cumberlandensis (ANSP 141, 142) and Microtus (Pedomys) sp. (ANSP 18907) from Port Kennedy Cave with other Pleistocene samples. Cumberland Cave data from van der Meulen (1978), Trout Cave 2 data from Pfaff (1990). Numbers for samples other than Port Kennedy Cave include the mean and observed range.

L a w’ W b C A/L W’/W B/W C/W Pitymys cumberlandensis Port Kennedy Cave ANSP 141 2.28 1.13 0.78 0.84 0.12 0.06 48.56 92.86 14.30 7.14 ANSP 18907 2.56 1.00 0.81 0.95 0.05 0.12 39.06 85.26 5.26 12.63

Cumberland Cave 2.54 48.2 92.6 16.2 15.1 (2.28-2.87) (44-53) (84-102) (6-27) (6-32) Trout Cave 2 2.62 47.5 91.5 14.1 15.4 (2.36-2.92) (42.2-53.1) (79-101.2) (7.3-24.1) (8.5-26.0)

M. (Pedomys) llanensis Trout Cave 2 3.18 47.2 84.8 19.8 21.0 (2.51-3.59) (39.5-52.1) (73.7-98.5) (9.9-30.6) (13.3-29.0)

M. (Pedomys) guildayi Cumberland Cave 2.58 45.7 80.4 26.0 20.4 (2.27-3.07) (42-50) (70-91) (13-36) (9-31)

M. (Pedomys) sp. Port Kennedy Cave ANSP 142 2.57 1.18 0.87 1.11 0.12 0.23 45.91 78.38 14.29 27.38

______complexity of the M3, with shallow and generally probably preclude the known populations of P. horizontal reentrants. What separates P. cumberlandensis from direct ancestry to P. pinetorum cumberlandensis M3s from those of P. pinetorum and and other modern and extinct pine voles in which these other known North American Pitymys is that the apices fields are mostly closed. Thus, P. cumberlandensis of T2-3 are directly opposite and the dentine fields of likely represents a sister species of the species ancestral these triangles are entirely confuent (Pfaff, 1990: fig. to later pine voles. The ancestral North American pine 8). Confluence of T2-3 in M3 was reported only in vole likely displayed some of the features of P. modern P. pinetorum parvulus from Florida (Arata, cumberlandensis, such as thick, undifferentiated 1965), which as we discussed above has the simplest enamel, unreduced m3 and a tendency in m1 for T4-5 M3 in P. pinetorum. As illustrated by Arata (1965, fig. to be closed, but it would also display closed or nearly 4D) M3s of P. pinetorum from the late Pleistocene closed T2-3 on M3 and closed or nearly closed T1-2 on Reddick 1A deposit of northcentral Florida m1 and m2. It is conceivable that this “missing link” is demonstrate a tendency towards confluent and opposite simply another population of P. cumberlandensis in T2-3, but were not as extreme in their morphology as which these features are expressed. In some support for the M3s of P. cumberlandensis or those from living this idea is the observation in the Trout Cave 2 Florida pine voles. These observations lead to the population of P. cumberlandensis that the dominant m1 following tentative conclusions: 1) the confluent T2-3 morphology is one in which T1-2 are closed rather than on a complex M3 as in P. cumberlandensis is likely a open (Pfaff, 1990: fig. 8). derived, rather than primitive condition; 2) the confluent M3 of extant P. pinetorum evolved relatively AGE OF THE PORT KENNEDY CAVE recently, during the late Pleistocene, and is convergent, ASSEMBLAGE rather than homologous, with the condition in P. cumberlandensis. Daeschler (1996) published a complete list of Another unusual dental feature of P. vertebrates from the Port Kennedy assemblage and cumberlandensis is the tendency in all lower molars for concluded that the site was late Irvingtonian in age, the dentine fields of T1-2 to be confluent (Figure 2). possibly comparable to Hamilton Cave in West The well-developed dentine field of T4 on M3, another Virginia (Repenning and Grady, 1988) and older than distinguishing feature of P. cumberlandensis, is also Cumberland Cave (Gidley and Gazin, 1938) in directly opposite, rather than alternating with, the Maryland and Trout Cave 2 (Pfaff, 1990) in West dentine field on T3. Taken collectively, the open and Virginia. Based on the rodents, Daeshler’s (1996) confluent dentine fields in both upper and lower molars general estimate of age for these early Pleistocene sites 18 MARTIN—PORT KENNEDY ARVICOLID RODENTS

TABLE 2. Microtus and Pitymys species from select midwestern and eastern early and middle Pleistocene assemblages. Subg, indet. = Subgenus indeterminate.

Approx. age (Ma) (2.0) (1.6-1.3) (0.64) (0.35)

Species Aries A Cumberland Cave, Port Kennedy Cave, Coleman 2A Hamilton Cave Trout Cave 2, Cudahy

Microtus (Allophaiomys) M. pliocaenicus X Microtus (Pedomys) M. guildayi X M. llanensis X Microtus (subg. indet.) M. paroperarius X X M. meadensis X Pitymys P. cumberlandensis X X P. aratai X ______from the northeastern United States seems reasonable, llanensis. Consequently, Cumberland Cave is likely but a different temporal order is likely. older than Trout Cave 2 and Port Kennedy Cave The combination of Ondatra annectens (= O. (assuming the M. cf. llanensis identification of ANSP hiatidens), Pitymys cumberlandensis and Microtus 142 is correct). (Pedomys) cf. llanensis suggests that the Port Kennedy The early Pleistocene Hamilton Cave arvicolids assemblage is referable to Martin’s (2003) Rodent from West Virginia need restudy. Repenning and Zone (RZ) 14, originally developed for the Great Plains Grady (1988) listed 30 m1s of Microtus region. RZ 13 includes a series of local faunas with the (Allophaiomys) pliocaenicus from the Cheetah Room archaic arvicolid Microtus pliocaenicus, and RZ 14 of Hamilton Cave, but none of the specimens were includes M. (Pedomys) llanensis (=M. ochrogaster illustrated. If this identification is confirmed, then the /llanensis of Martin et al. [2008]) as well as M. Cheetah Room material would be assignable to RZ 13. paroperarius, also found at Trout Cave 2 (Pfaff, 1990) However, Repenning and Grady (1988) also mentioned and Cumberland Cave (van der Meulen, 1978). that some of the m1s display a Microtus (Pedomys) Midwestern and western assemblages of RZ 14 also guildayi morphology. In their assessment of fossil bog usually include Microtus meadensis (sometimes lemmings and muskrats from Hamilton Cave, Martin et referred to Terricola), but the absence of this species in al. (2003a, 2009) concluded that the Cheetah Room the eastern United States is likely due to ecological and Smilodon Room rodent collections from Hamilton rather than temporal factors.Thus, Trout Cave 2 and Cave were deposited between 1.0-1.6 Ma, more likely Port Kennedy Cave may be approximately coeval toward the older end of the estimate (1.3-1.45 Ma). eastern equivalents of the Cudahy and Sunbrite local Repenning and Grady (1988) also identified Microtus faunas of Kansas (Martin et al., 2003b). The Cudahy deceitensis (as Lasiopodomys deceitensis), (cf.) M. local fauna lies directly beneath the Lava Creek B ash paroperarius, and Pitymys hibbardi from the Cheetah at 0.64 Ma, but we do not know either the earliest or Room. There is a superficial resemblance between latest records of the arvicolids rcorded from this site. some of the specimens illustrated by Repenning and Only Microtus pliocaenicus has been recorded beneath Grady (1988) to those of M. deceitensis and M. morlani the Cerro Toledo B ash (about 1.35 Ma) in the Meade from the Yukon (Storer, 2003, 2004), but none of the Basin of Kansas (Martin et al., 2008) and the extant M. Hamilton Cave m1s express the short, wide and deep ochrogaster is found in Great Plains assemblages of lingual triangles that characterize the Alaskan species. the late Pleistocene. RZ 14, therefore, conceivably Instead, the Hamilton Cave specimens display could extend from about 1.35-0.30 Ma. For now, it relatively longer and narrower triangles with more would not be unreasonable to provide an approximate acute apices, more like M. paroperarius (Hibbard, correlation of the Trout Cave 2 and Port Kennedy Cave 1944). Additionally, M. deceitensis molars display assemblages with Cudahy, at about 0.64 Ma. undifferentiated enamel, and if the illustrations are Microtus (Pedomys) guildayi from Cumberland accurate, the specimens illustrated by Repenning and Cave displays a simpler dentition than that of M. Grady (1988, fig. 4) display positive differentiation on llanensis from Trout Cave 2 and Cudahy (van der many edges, another similarity to M. paroperarius. In Meulen, 1978; Pfaff, 1990) and may be directly any case, allocation to the genus Lasiopodomys for the ancestral to at least the eastern populations of M. Hamilton Cave specimens is certainly unwarranted at PALUDICOLA, VOL. 9, NO. 1, 2012 19

this time. Without further study of the Hamilton Cave Ted Daeschler at the Philadelphia Academy of material all we can state with certainty about the Sciences. Comparative fossil material used in this relatively advanced Microtus from the Cheetah Room study from the Meade Basin of Kansas was collected is that it represents a species of Microtus with an m1 in through National Science Foundation grant EAR- which T4 is relatively reduced and T6 is absent or 0207582. The detailed review of this paper by R. minimally expressed, both of which are features of M. Zakrzewski is much appreciated. paroperarius and the extant M. oeconomus group. Pitymys hibbardi from the Williston 3A local LITERATURE CITED fauna of Florida is a junior synonym of P. pinetorum (Martin, 1995), and differs in numerous ways from the Arata, A. A. 1965. Taxonomic status of the pine vole in Pitymys specimens from Hamilton Cave, three of Florida. Journal of Mammalogy 46:87-94. which were illustrated by Repenning and Grady as P. Cope, E. D. 1871. Preliminary report on the Vertebrata hibbardi (1988, fig. 5). Illustrations B (USNM discovered in the Port Kennedy Bone Cave. 264313) and C (USNM 264314) from their fig. 4 are Proceedings American Philosophical Society very similar to specimens of P. cumberlandensis van 12:73-102. der Meulen (1978), but illustration D (USNM 25431) Daeschler, E. B. 1996. Selective mortality of has such a simple anteroconid complex that it is mastodons (Mammut americanum) from the unclear if this specimen is a Pitymys or Microtus Port Kennedy Cave (Peistocene: (Allophaiomys); Repenning and Grady (1988) mention Irvingtonian), Montgomery County, Penn- this possibility in the legend to their fig. 5. Pitymys sylvania. Pp. 83-96 in K. M. Stewart and K. cumberlandensis thus appears to be a likely member of L. Seymour (eds.), Palaeoecology and the Hamilton Cave assemblages, but the sample needs Palaeoenvironments of Late Cenozoic to be compared with those from Cumberland Cave and Mammmals: Tributes to the Career of C. S. Trout Cave 2. Based on published information, it (Rufus) Churcher. University of Toronto seems likely that a relatively primitive Microtus near Press, Canada. M. guildayi is present at Hamilton Cave, which Gidley, J. W. and C. L. Gazin. 1938. The Pleistocene suggests an age for Hamilton Cave at least as old as vertebrate fauna of Cumberland Cave, Cumberland Cave, and likely older than Port Kennedy Maryland. Bulletin U. S. National Museum Cave and Trout Cave 2. 1:1-99. The Coleman 2A l.f. of Florida, also allocated to Hibbard, C. W. 1944. Stratigraphy and veretebrate RZ 14 by Martin (2003), probably was deposited at the paleontology of Pleistocene deposits of younger end of RZ 14, as the Coleman 2A Pitymys southwestern Kansas. Bulletin Geological aratai (Martin, 1974) displays positive enamel Society of America 55:718-744. differentiation and a reduced m3 (Martin, 1995), both Hibbard, C. W. 1955. Notes on the microtine rodents advanced characters relative to P. cumberlandensis. from the Port Kennedy Cave deposit. The records above are summarized in Table 2. Proceedings of the Academy of Natural In conclusion, due to the absence of holotypes Sciences of Philadelphia 107:87-97. and most paratypes, all of Cope’s (1871) Microtus-like Martin, R. A. 1974. Fossil mammals of the Coleman species from Port Kennedy Cave are declared nomina IIA fauna, Sumter County. Pp 35-99 in dubia. Two remaining specimens from Port Kennedy S. D. Webb (ed.), Pleistocene Mammals of Cave can be assigned with certainty to Pitymys Florida. University of Florida Press, cumberlandensis van der Meulen (1978) and one Gainesville. tentatively to M. (Pedomys) llanensis (Hibbard, 1944). Martin, R. A. 1987. Notes on the classification and Based on these specimens, Port Kennedy Cave is evolution of some North American Microtus. allocated to the early Pleistocene in RZ 14, and is Journal of Vertebrate Paleontology 7:270-283. therefore younger than other eastern early Pleistocene Martin, R. A. 1989. Arvicolid rodents of the early assemblages such as Hamilton Cave and Cumberland Pleistocene Java local fauna from north- Cave. Port Kennedy Cave is considered to be roughly central South Dakota. Journal of Vertebrate coeval with Trout Cave 2 and, in the midwest, with Paleontology 9:438-450. Cudahy from the Meade Basin of Kansas at about 0.64 Martin, R. A. 1991. Evolutionary relationships and Ma. biogeography of late Pleistocene prairie voles from the eastern United States. Pp. 251-260 ACKNOWLEDGEMENTS in J. R. Pordue, W. E. Klippel and B. W. Styles (eds.), Beamers, Bobwhites and Blue- Three Port Kennedy Cave arvicolid specimens Points: Tributes to the Career of Paul W. from Port Kennedy Cave were provided for study by 20 MARTIN—PORT KENNEDY ARVICOLID RODENTS

Parmalee. Illinois State Museum Scientific Martin, R. A., F. Marcolini, and F. Grady. 2009. The Papers 23, Springfield. early Pleistocene Hamilton Cave muskrats and Martin, R. A. 1995. A new middle Pleistocene species a review of muskrat size change through the of Microtus (Pedomys) from the southern late Neogene. Paludicola 7:61-66. United States, with comments on the tax- Martin, R. A., P. Peláez-Campomanes, and J. G. Honey onomy and early evolution of Pedomys and 2012. Preliminary study of rodents from the Pitymys in North America. Journal of Golliher B assemblage of Meade County, Vertebrate Paleontology 15:171-186. Kansas, USA indicates an intense cold period Martin, R. A. 2003. Biochronology of latest Miocene near the end of the Pleistocene. Palaeontologia through Pleistocene arvicolid rodents from the Electronica, article 14.3.35A:1-13. Central Great Plains of North America. Pp. Meulen, A. J. van der. 1978. Microtus and Pitymys 373-383 in N. López-Martinez, P. Peláez- (Arvicolidae) from Cumberland Cave, Campomanes and M. Hernández Fernández Maryland, with a comparison of some New (eds.), Surrounding Fossil Mammals: Dating, and Old World species. Annals of Carnegie Evolution and Paleoenvironment. Colloquios Museum 47:101-145. de Paleontología, Volumen Extraoraordinario Pfaff, K. S. 1990. Irvingtonian Microtus and Pitymys 1 en Honor de Dr. Remmert Daams. Madrid. (Mammalia, Rodentia, Cricetiade) from Trout Martin, R. A., A. C. Bonk, and P. Peláez-Campomanes. Cave No. 2, West Virginia. Annals of 1998. Variation in enamel differentiation of Carnegie Museum 59:105-134. Microtus molars from the Java local fauna, Repenning, C. A. and F. Grady. 1988. The microtine Walworth County, South Dakota. Paludicola rodents of the Cheetah Room fauna, Hamilton 2:74-77. Cave, West Virginia, and the spontaneous Martin, R. A., L. Duobinis-Gray, and C. P. Crockett. origin of Synaptomys. U. S. Geological 2003a. A new species of early Pleistocene Survey Bulletin 1853:1-32. Synaptomys (Mammalia, Rodentia) from Storer, J. E. 2003. The eastern Beringian vole Microtus Florida and its relevance to southern bog deceitensis (Rodentia, Muridae, Arvicolinae) lemming origins. Journal of Vertebrate in late Pliocene and early Pleistocene faunas Paleontology 23:917-936. of Alaska and Yukon. Quaternary Research Martin, R. A., R. T. Hurt, J. G. Honey, and P. Peláez- 60:84-93. Campomanes. 2003b. Late Pliocene and Storer, J. E. 2004. A middle Pleistocene (late Pleistocene rodents from the northern Irvingtonian) mammalian fauna from Thistle Borchers Badlands (Meade County, Kansas). Creek, Klondike Goldfields region of Yukon With comments on the Blancan-Irvingtonian Territory, Canada. Paludicola 4:137-150. boundary in the Meade Basin. Journal of Paleontology 77:985-1001. Martin, R. A., P. Peláez-Campomanes, J. G. Honey, D. L. Fox, R. J. Zakrzewski, L. B. Albright, E. H. Lindsay, N. D. Opdyke and H. T. Goodwin. 2008. Rodent community change at the Pliocene-Pleistocene transition in southwestern Kansas and identification of the Microtus immigration event on the Central Great Plains. Palaeogeography, Palaeo- climatology, Palaeoecology 267:196-207.